|
Abstract
|
Correction to: International Journal of Biometeorology
https://doi.org/10.1007/s00484-022-02269-3
Following publication, we discovered a data processing error that misassigned North American Land Data Assimilation System 2 (NLDAS-2)-derived temperature index values, affecting a subset of locations over the assessed time period. Upon correcting this error and repeating the analysis, the findings remained fundamentally unchanged, although the partial misassignment had biased risk estimates towards the null. The corrected odds ratios increased marginally for heat index and WBGT indices, and by a larger magnitude for ambient temperature and minimum/early morning indices, relative to case-day median index values, which also increased upon re-analysis. The shapes of the response curves for the multiple indices remained consistent with the exceptions of minimum temperature, morning (0600 local) temperature, and morning WBGT, presented in the electronic supplementary material.
The abstract findings are amended to "Responses were positive, monotonic, and exponential in nature, except for maximum daily WBGT, which showed decreasing risk for the highest heat category days. The risk for heat stress illness on a day with a maximum WBGT of 32.2 °C (90.0 °F) was 2.08 (95% CI, 1.93-2.23) times greater than on a day with a maximum WBGT of 29.9 °C (85.8 °F) (prior: OR = 1.93, relative to 28.6 °C). The risk was 3.15 (2.92-3.41) times greater on days with a maximum heat index of 40.6 °C (105 °F) compared to 34.7 °C (94.4 °F) (prior: OR = 2.53, relative to 32.8 °C)."
Numerous tables and figures are updated and are available at https://doi.org/10.1007/s00484-022-02269-3
Language: en |